Field of the Invention
Exemplary embodiments of the present invention relate to improved traffic sign recognition of a driver assistant system capable of classifying circular shapes indicating circles of objects in a front image acquired by a front looking camera equipped in a vehicle and recognizing the classified circles as a circular traffic sign when the classified circles are associated with a white inner circle.
Description of the Related Art
In recent years, a driver assistant system may automatically provide traffic information to a driver in real time to promote driving convenience and accident prevention.
Among the driving assistant systems, a traffic sign recognition (TSR) system is a system which uses a camera equipped in a vehicle to recognize a front traffic sign to inform a driver of a current limit speed, a lowest speed, and other road information.
In particular, accurately recognizing the traffic sign using an image is very important in performance of the TSR system.
An advanced driver assistance system (ADAS) image recognition apparatus processes an image input in a digital form through the camera based on a numerical analysis technique, and therefore performance of the ADAS image recognition apparatus may largely rely on characteristics of the input image.
A reflective traffic sign goes through a process of allowing sunlight reflected from the traffic sign during the day and streetlights and headlights of a self vehicle and other vehicles mainly reflected from the traffic sign during the night to be incident on the camera of the ADAS image recognition apparatus and then converting the reflected sunlight and streetlights and headlights into digital images, and therefore is greatly affected by brightness or directions of surrounding light sources.
On the other hand, an emissive traffic sign is not affected by surrounding light sources since light generated from its own light source is incident on the camera of the image recognition apparatus, regardless of other light sources. Due to the difference in characteristics as described above, it is difficult to recognize two kinds of traffic signs by photographing the traffic signs using the same camera.